Prison fights need data referees

Dr Eric Crampton
Insights Newsletter
31 July, 2015

When the government moved to privatise Mount Eden prison, I was a sceptic. There can be very good arguments for outsourcing service provision or privatisation in all kinds of sectors, but prisons were about the last on my list.

Why? Because it can be pretty tough to write contracts that do not give prison operators an incentive to cut costs by worsening conditions for prisoners in areas that are not monitored in the contract. If innovation does not matter much, arguments for outsourcing are weaker. Keeping people locked in cells did not seem a sector in need of entrepreneurial disruption. And American experience with private prisons, where judges would take kickbacks from prison operators for throwing kids in private cells, also gave me pause.

Shortly afterwards, the Corrections Association launched a harsh critique against Serco. Why? Because Serco was being too nice to the prisoners, choosing to keep prisoners in line with better conditions rather than with more guards. The economics literature suggests that harsher prisons make recidivism more likely. Since Serco stands to gain substantial bonuses if reoffending rates drop, they have some incentive to get this right. Innovation in reducing recidivism is worthwhile.

Over the past two weeks, the news has shifted from a fight club at Serco’s Mount Eden remand facility to poor conditions across the sector – including in government-run prisons. Unfortunately, the data is not that great. The latest Prison Performance Table, for the year ending March 2015, had Mount Eden at the top of the heap, but it is not easy to tell how you would compare a remand facility with other prisons. And while there is background data on Serco’s performance, because its contracted payments require it, we just do not have the same level of detail on public prisons. Finally, the anecdotes coming out over the past couple of weeks make some of the data claims seem implausible: were there really only 0.74 serious assaults per 100 prisoners at Mount Eden?

The lack of good comparative data lets ideologically driven stories dominate. We could follow Voltaire’s Candide and ask, if Mount Eden is the best of all possible prisons, what hells be all the others? Until the government collects, verifies, and publishes better statistics, it will not be easy to tell.

Stay in the loop: Subscribe to updates